Insulin-like growth factors (IGF's) are mitogenic peptides that regulate embryonic development, post-natal growth and cellular differentiation in vertebrates. The functions of mature IGF peptides have been extensively studied in various in vitro and in vivo systems. IGF's, including IGF-I and IGF-II, are among the members of a family of structurally and evolutionarily related peptides that also include insulin and relaxins. Like many hormones, IGF's are initially translated as pre-pro-peptides that undergo post-translational processing to result in the mature peptides.
The mature form of mammalian IGF-I is a basic protein of 7.5-kDa. The pre-pro-peptides of the mammalian IGF-I consist of an amino-terminal signal peptide, followed by the mature peptide with B, C, A and D domains, and a carboxyl-terminal E domain (See FIG. 1A for a schematic representation). The signal peptide at the amino-terminal end and the E-domain peptide at the carboxy-terminal end of the pre-pro-peptide are proteolytically cleaved from the peptide to result in the mature, biochemically active species.
To date, multiple forms of pro-IGF-I have been identified in species from fish to mammals (Shamblott, Chen, Mol Mar Biol Biotechnol. 2: 351–61, 1993; Rotwein, Proc. Natl. Acad. Sci USA, 83:77–81, 1986). In humans, three alternative spliced isoforms of pro-IGF-I (pro-IGF-I-a pro-IGF-I-b and pro-IGF-I-c) have been reported (Rotwein, Proc. Natl. Acad. Sci USA, 1986; Rotwein, et al., J. Biol. Chem., 261: 4828–32, 1986; Chew, et al., Endocrinology, 136: 1939–44, 1995). These three pro-IGF-I isoforms differ only in the carboxyl-terminal E-domain regions that are normally removed in vivo from the mature IGF-I. The E-domains of pro-IGF-I-a, pro-IGF-I-b and pro-IGF-I-c contain 35, 77 and 40 amino acid residues, respectively. The first 15 amino acid residues at the N-terminus of E-domains (referred to as the common region) share identical sequences. The amino acid sequences following the common region vary between the three isoforms of human pro-IGF-I (see FIG. 1B).
Similar diversity of pro-IGF-I E-domains is also found in rainbow trout (Oncorhynchus mykis), where four different isoforms have been identified, designated for consistent reference herein as pro-IGF-I Ea-1, Ea-2, Ea-3 and Ea-4 (Shamblott, Chen, Mol. Mar.Biol. Biotech., 1993). Nucleotide sequence comparison of the four size forms of rainbow trout IGF-I mRNAs is consistent with the above observations concerning the Ea peptides in that the size differences among these mRNA species are due to insertions or deletions in the E domain regions of the molecules (See FIGS. 1A and 1B). The predicted amino acid residues of the common region of the four Ea peptides share identical sequences among themselves, as well as with pro-IGF-I E-peptides of human, mouse, and rat species (See FIG. 1B). The presence of the C-terminal 20 amino acid residues, sharing 70% identity with their human counterparts, identifies them as a-type E-peptides. The Ea-1 peptide of the rainbow trout (rt) pro-IGF-I (SEQ ID NO: 5) is a polypeptide of 35 amino acid residues, comprising the first 15 and the last 20 amino acid residues. Ea-2 (SEQ ID NO: 4) and Ea-3 (SEQ ID NO: 3) peptides differ from Ea-1 by virtue of either a 12- or 27-amino acid residue insertion between the first and last segments of the Ea-1-peptide sequence, respectively (see FIG. 1B). The Ea-4 peptide (SEQ ID NO: 2) contains both insertions. The predicted numbers of amino acid residues in each E-peptide are, thus, 35 (SEQ ID NO: 5), 47 (SEQ ID NO: 4), 62 (SEQ ID NO: 3) and 74 (SEQ ID NO: 2), respectively. There has not been any report on the presence of b-type IGF-I mRNA in rainbow trout (Shamblott and Chen, 1993).
FIG. 1B shows the amino acid sequences of the human Eb peptide (hEb) (SEQ ID NO: 1) and the trout Ea peptides. Despite not having complete homology at the primary level, preliminary studies (unpublished data of this laboratory) indicate that hEb and trout Ea-4 peptide have very similar tertiary structures, particularly in the amino-terminal region containing the common sequences, and can compete effectively for binding to cell receptors specific to E-domain peptides.
Despite the presence of multiple E-domain variants, assigning biological function to the IGF E-domains has been elusive. Proteolytic processing of the pro-IGF's, resulting in the cleavage of E-domains from IGF's, is believed to be similar to the cleavage of the C-peptide of proinsulin (Foyt, et al., Insulin-Like Growth Factors: Molecular and Cellular Aspects, pp1–16. Boca Raton: CRC press, 1991). In the past, it was generally accepted that E-domains, like the C-peptide of pro-insulin, possess little or no biological activity other than their potential roles in the biosynthesis of mature IGF. The C-peptide of pro-insulin is believed to have an essential function in the biosynthesis of insulin in linking the A and B chains in a manner that allows correct folding and inter-chain disulfide bond formation. In spite of the earlier reports indicating certain physiological effects of the insulin C-peptide (Johansson, et al., Diabetologia, 35: 121–28, 1992; Johansson, et al., Diabetologia, 35: 1151–58, 1992; Johansson, et al., J. Clin. Endo. Metab., 77: 976–81, 1993), it has not been widely accepted until recently. The C-peptide has now been shown to have many beneficial effects on various abnormalities in diabetic animal models and patients (Ido, et al., Science, 277: 563–66, 1997; Forst, et al., J. Clin. Invest. 101: 2036–41, 1998; Sjoquist, et al., Kidney Int., 54: 758–64, 1998). Moreover, recent studies further demonstrated specific binding of C-peptide to cell surfaces in a manner that suggests the presence of G-protein-coupled membrane receptors (Rigler, et al., Proc. Natl. Acad. Sci USA, 96: 13318–23, 1999). It is now thought that C-peptide may thereby stimulate specific intracellular signal transduction leading to the biological activities of C-peptide (Wahren, et al., Am. J. Physiol. Endo. Metab. 278: E759–68, 2000; Kitamura, et al., Biochem J., 355: 123–29, 2001).
Tian et al. (1999) have recently reported that recombinant rainbow trout Ea-2-, Ea-3- and Ea-4-peptides possess mitogenic activity in several non-transformed cell lines, including NIH 3T3 cells and caprine mammary epithelium cells (CMEC) (Panschenko et al., 1997). Since trout Ea-2-and Ea-4-peptide contains a signal motif for peptidyl C-terminal amidation (Shamblott and Chen, 1993; Barr, 1991), and a bipartite consensus nuclear localization sequence is also present in Ea-4-peptide (Shamblott and Chen, 1993; Dingwall and Laskey, 1991), the present inventors have concluded that these peptides potentially possess other novel biological activities. Thus, the present inventors demonstrate that novel biological activities are associated with both the Ea peptides of the rainbow trout pro-IGF-I, and with the human Eb peptide.